Where the number comes from
Air at ordinary pressures behaves closely enough like an ideal gas that its density is 39.67 divided by its absolute temperature in degrees Rankine, in pounds per cubic foot. At 60 degrees Fahrenheit that gives 0.0763, which is the figure every ventilation calculation on this site uses. Take the density of the outdoor air, subtract the density of the flue gas, multiply by the vertical height, and you have a pressure difference in pounds per square foot. One inch of water column is 5.2027 pounds per square foot, so dividing by that puts the answer in the unit a manometer reads.
On the numbers the form opens with: outdoor air at 30 degrees is 489.67 Rankine and weighs 0.08101 lb per cubic foot. Flue gas at 350 degrees is 809.67 Rankine and weighs 0.04900. The difference is 0.03202 lb per cubic foot, and over 20 feet that is 0.6404 lb per square foot, or 0.1231 inches of water column. In metric that is 30.66 pascals. Every term in that is visible and none of it is fitted to anything.
Three levers, and only one of them is cheap
Height is linear: every foot adds the same amount, 0.00615 in w.c. per foot on these conditions. Flue temperature is not linear, because it enters as a reciprocal, so the first hundred degrees buys far more than the fifth hundred. Outdoor temperature runs the other way and is not yours to set.
| Change from the default case | Draft | Difference |
|---|---|---|
| As entered: 20 ft, 350°F gas, 30°F out | 0.1231 | — |
| Gas 50°F hotter | 0.1340 | +0.0110 |
| Gas 50°F cooler | 0.1107 | −0.0124 |
| 10°F colder outside | 0.1296 | +0.0065 |
| 20°F warmer outside | 0.1109 | −0.0122 |
| Five more feet of chimney | 0.1539 | +0.0308 |
| Same chimney at 5,000 ft elevation | 0.1024 | −0.0207 |
The elevation row surprises people who move a stove from the coast to the mountains and find it will not draw the way it used to. At 5,000 feet the barometric pressure is about 83.2 percent of sea level, both air columns are proportionally lighter, and so is the difference between them. Nothing about the chimney changed.
The case where the number goes negative
Put the flue gas temperature below the outdoor temperature and the arithmetic returns a negative draft, and this is not a contrived input. An exterior masonry chimney on the cold side of a house holds a large mass of masonry that has been losing heat all night. In the morning that flue can be colder than the outside air, and it is certainly colder than the house. The heavy column of cold air in the flue sits there, and if anything is pulling on the house the flow goes down the chimney and into the room.
That is the condition where a first match fills a room with smoke, and it is the same condition, without a fire, in which an unused flue delivers a steady trickle of cold air and whatever is in it. The hazards are spillage of combustion products, and carbon monoxide, which has no smell and does not announce itself. This page names that and gives no procedure for dealing with it, because there is no procedure that is safe to publish for a chimney nobody has looked at. That is a conversation with a certified sweep.
What the theoretical figure is not
It is not a prediction of what a manometer will read. Real draft at the appliance is the theoretical figure minus everything the gas has to push through: the connector, every elbow, the flue itself, the cap, and the appliance. It is also minus, or occasionally plus, whatever the house is doing. A tight house with a 600 CFM range hood running can go negative by more than the entire number above, and then the flue is simply the largest hole available for makeup air. The makeup air calculator puts a size on that effect, and the air sealing payback calculator is where the same problem is created accidentally by good intentions.
If you want the real number, somebody with a manometer takes it at the connector with the appliance at operating temperature and the house in its normal running state, exhaust fans and all. This calculator gives you the ceiling that measurement has to fit under, and the sensitivity to know which variable is worth arguing about.
Questions people ask
My chimney draws badly on mild days. Is that a fault?
It is what the arithmetic predicts. Draft is driven by the temperature difference between the flue gas and the outside air, so a 55 degree afternoon in October produces far less of it than a 10 degree night in January. On the default case here, going from 30 degrees outside to 50 degrees takes the theoretical draft from 0.1231 to 0.1109 in w.c., and the drop steepens as the fire itself burns cooler in mild weather. Whether anything is actually wrong with the chimney is a question for a sweep who can look at it, and this page has no opinion on it.
Would a taller chimney fix a draft problem?
The arithmetic says height adds draft linearly, so five more feet on the default case is worth 0.0308 in w.c., a quarter more than it had. Whether height is the actual constraint, whether the chimney can carry more, what height is permitted relative to the roof, and whether the problem is really resistance or house pressure rather than height are all things this page cannot tell you. Chimney height above a roof is set by the code your jurisdiction adopted and by the appliance listing, and adding to a chimney is work for a professional.
Why does an insulated liner draw better than a bare masonry flue?
Because the term that matters is the average gas temperature over the whole height, not the temperature at the collar. Gas leaving a stove at 400 degrees can arrive at the top of a cold exterior masonry chimney under 200, and the average across that run drives the buoyancy. Anything that keeps the gas hot on the way up keeps the draft up with it, and the same effect keeps the flue above the temperature at which unburned gas condenses on the wall. What liner is appropriate for a given chimney and appliance is a question for the manufacturer and a certified sweep.
What flue gas temperature should I enter?
The average over the height, which is lower than anything you can read at the connector. If you have a probe thermometer in the connector, that reading is the hot end of the range. A reasonable way to use the field is to bracket it: run it at the connector reading and again at half of it, and treat the answer as a band. On the default 20 ft chimney, 350 degrees gives 0.1231 and 200 degrees gives 0.0803, so the honest answer spans about a third.
Does the calculator tell me if my draft is adequate?
No, and it deliberately will not. Adequacy is defined by the installation instructions for the specific appliance, verified by a measurement at the appliance, in the house, with the house running the way it normally runs. The optional fields let you put a manufacturer figure and a manometer reading next to the theoretical number so you can see the shape of the gap, but the page draws no conclusion from that and neither should you without the person who took the reading.